Identifying impact coverage gaps through hybrid LCA-IO

A case study of German car manufacturing under different transition scenarios.

Territorial greenhouse gas (GHG) emissions of Germany have declined by over 30 percent compared to 1990 levels. The majority of this decline is due to a shift in the industrial landscape and a small contribution through renewable energy production. However, more profound changes in the industrial landscape as well as in society are required for Germany to meet its climate goals. These goals are set out in the climate action plan and entail 55%  GHG emissions reduction by 2030 and 95% by 2050. The German Environmental Agency has developed 6 different scenarios in their Resource-efficient Pathways towards Greenhouse-Gas-Neutrality (RESCUE) report, based on different levels of ambition towards achieving GHG neutrality. Given that the transport sector is responsible for nearly 20 percent of the national emissions in Germany, it is therefore not surprising, that in all RESCUE scenarios the shift from combustion engine passenger vehicles to electric passenger vehicles contributes significantly towards GHG neutrality. 

Although life cycle impacts of different types of passenger vehicles is a well studied topic, the impacts of their production phase on an economy wide scale has not been in the focus. It is however highly relevant, as considerable burden-shifting may occur due to the energy and material-intensive production of vehicles with new drive technologies. The complete supply chain impacts of passenger vehicle manufacturing therefore needs to be taken into account when designing pathways to meet national climate targets. 

While process-based life cycle assessment  (LCA) databases such as ecoinvent are continuously being updated to include more and updated processes, significant data gaps still limit the system completeness which can lead to underestimation of the environmental impacts due to truncation. Environmentally extended input-output (EEIO) databases, on the other hand, lack the amount of detail required to differentiate between different vehicle types in transition pathways such as the RESCUE scenarios. Moreover, the inclusion of capital creation can have a significant impact on the environmental burdens associated with a supply chain, something which is highlighted by the recent work on the inclusion of capital formation into EEIO analysis.  However, these studies also emphasise the non-trivial nature of capital inclusion into EEIO databases, whereas capital creation is naturally included in attributional process-based LCA. 

Therefore, hybrid LCA-IO analysis remains an important tool to study complete supply chain footprints, such as those of passenger vehicle manufacturing in scenarios from RESCUE. 

Hybrid LCA-IO analysis, however, is in itself not a straight forward process either, and all versions of hybrid analysis have multiple challenges that lead to added uncertainty in the final footprints.

In this work we compare the performance of a hybrid LCA-IO analysis to more traditional LCA and IO analyses in the case of environmental impacts along the supply chains in the production phase of passenger vehicles under different scenarios from the RESCUE study. For this study we use the EXIOBASE environmentally-extended IO database and the ecoinvent LCA process inventory, the BACI trade data base, as well as the scenario data from the RESCUE report by the German Environmental Agency.  As emerging technologies are not yet captured in either EXIOBASE nor ecoinvent, we will add own foreground data to the hybrid system, such as advances in battery energy density and recycling efficiencies. Moreover, we use a novel approach to capture the added uncertainty that arises from the hybridisation process as well as the uncertainty in the process data part of the system. This approach enables us to assess not only the differences of total supply chain footprints between the various hybridisation methodologies, but also the trade-offs in added uncertainty of the final results.